Sunday 3 October 2010

Evolving 'Scientists' into 'Science-twists'?

First of all, it is worth defining the First of all, it is worth defining the terms 'scientist' and science-twist' in the context of this article. The dictionary definition of the scientist is 'the person with an advance knowledge of one or more sciences'. Here, the term 'science-twist' is defined as 'the person with an advance knowledge of various science disciplines without (or with) an intention of compromising the quality of the results for achieving desired outcomes'.

The objectives of this article is to look at the current research practices in the light of both the above defined terms, and to discuss the topic 'whether today's scientists are turning into science-twists?'.

As discussed in the article below, pecuiliar combination of output oriented workload on the academics and contention for publishing large number of articles to prove their scientific productivty is a normal part of today's academic life. This also lead sometimes in research frauds by twisting the research data. This may not be completely true in cases when scientists working in research Institutes are only assigned research work though they may have their own concerns in terms of attracting funding for carrying out their research. However, this is nearly true in today's very competitive academic environment where a huge work load, in terms of administrative, management and teaching activities, are implicitly part of the job and large expectations for bringing in research projects and publishing in top ranked journals are generally kept from the academics. A recent article by Fanelli (2010) looked into this matter and concluded based on the the USA data that competitive academic environments increase not only scientists' productivity but also their bias; the same phenomenon might be observed in other countries where academic competition and pressures to publish are high.

A study by Bjork et al. (2006) reported that there were about 27,750 peer-reviewed scientific journals worldwide in 2006 that published over 1,35,000 articles. This figure has grown substantially over the last recent years. This is just an indication about the extent to which research is being published. The good aspect of this is an excellent dissimination of the research to the scientific community worldwide which is far bettter than the old days. However, the negative aspect is many of these articles may not even meet required basic standards and just be giving surfacial scientific information. Moreover, the new trend has emerged in publishing research articles, which is publishing research results into small pieces to increase the quantity of the articles published by an individual. However, one of the reasons for this is also the word limits imposed by the journal. Unlike most of today's research articles, old articles may be a bit longer with a detailed methodology which allows the users to reproduce the results or the data presented in them.

Another trend which have emerged in recent years is a unsatisfying hunger of research grants in academics. A general perception is that more the research fundings one have (no matter in what research area!), better he/she is having a chance for academic growth. A number of academics spend nearly most of their available time in writing grants and managing the already going on research projects. In another words, many of us are turning into 'research managers' rather than the 'research scientists' because of rarely doing research ourselves; this has given birth to a inverse-learning trend i.e. 'learing through sub-ordinate research staff rather than the other way around'. In fact, self-research is also not practically possible given the commitments to various activities (other than core research) that can probably end up in compromising the quality of research output.

There is a constant pressure on today's scientists to get positive publishable results and win research fundings. The current situation is not like it used to be years ago when scientists had ample of time and flexibility for publishing selective high quality research on their own will. To survive in highly competitive academic environment and to achieve given (or self set out) research targets, there is a less margin of error in today's academic environment for an individual to get negative unpublishible results which may also increase the chances of derialing from the original path and in turn affecting the quality of the research output.

There is no mantra to avoid this situtation. However, self periodic re-evaluation for maintaining a good balance between the quality and the quanitity of research, and choosing the best of the available options could help up to avoid the situations described above up to some extent.

Regards,
Prashant

Note: The thoughts presented in this article are solely of the author and do not represent any organisation, group or any other individual.

Saturday 17 July 2010

Measuring your scientific productivity and impact on science?

Academics, researchers or the scientists work hard for boosting their research and disseminate it through various channels such as conference presentations and articles, technical reports or journal publications. A question can always be raised that 'how to evaluate the individual's productivity and the impact on science or scientific community'. Whether this should relate with the number of articles he published or the quality of these articles? The one widely accepted measure, though opposed by many, is the hirch factor (h-factor). It is not perfect index but takes into account both the quality and the quantity of individual's publications.

What is h-factor?
The h-factor was first described by a physics Professor, J.E. Hirsch, in a PNAS article in 2005. He defined this as 'a scientist has index h if h of his or her Np papers have at least h citations each and the other (Np-h) papers have <=h citations each'. The aim of this article is not to describe the h-factor in detail but to present views on its application in real life. Detailed explanation of h-factor can be seen in the source article. The h-factor can be computed by using SCOPUS, ISI Web of Knowledge or any other tools such as scHolar index.

How does this relate your scientific productivity and impact?
The citation to your articles is the key for a better h-factor. For example, if you publish 250 articles and one of your article gets cited 250 times or more but among the remaining ones, for instance, only 10 other articles get cited 10 or less times; the h-factor in this case will be only 10. This example explains that if one has to increase his/her h-factor its crucial that the articles published by the individual gets continuous citation.

The citation can be directly related with the quality of the research presented in an article. However, there are contrasting thoughts about this. For example, in some cases you may find that despite having excellent science in an article it does not get citations. There could be several reasons for this, such as the novelty of research and its real-world application, dissemination media (i.e. the quality, popularity and the way of an journal to promote research), the name (and fame) of an individual scientists in a particular research area, selection of an appropriate journal for publication and the way of presenting the research in the article in a fresh, interesting and attracting manner, etc.

Is this fair to compare the scientists working in different disciplines using h-factor?
My answer to this question is 'no'. The reason is that articles in some areas are not very well cited as opposed to others; see a lost of journal impact factors here. For example, the most cited journal in Environmental Sciences and Engineering is having an impact factor of about 5. This is much lower if we consider a example of structural engineering in contrast to substantially higher impact factor journals in neuroscience (about 178). It also means that if one is working in an area of high impact factored journals and manage to publish frequently can have a high h-factor.

What 'numbers' are impressive in Engineering?
My personal view is that if you have a h-index less than 15, your impact could be considered 'average'. This could be 'good' when h-index is between 15 and 25, 'very good' between 25 and 35, excellent between 35 and 50, and outstanding over 50. However, I do not have any strong support to defend this classification except saying that I know few people in the area of Engineering that have h-factor just over 50; they are key people in their fields with outstanding individual contributions.

Concluding remarks
It is probably not the best way to measure individuals productivity and impact but we currently do not seem to have a better index. This number could be good to mention in short biographies or applications that are made for jobs or research grants. The h-factor can not grow over night, it requires time to get your articles cited and consistet efforts from your part to include better science in your articles. However, the important thing which I feel is doing the good science in your articles rather than aiming first for h-factor and then writing the articles.

I hope you will find this article interesting and useful. Please feel free to comment on it, if you have any!

Regards,
Prashant

Wednesday 9 June 2010

Tips for writing technical reports (BEng/MEng/MSc/EngD/D.Phil/PhD)

It is crucial to present your work in a concise, effective and impressive way. Failing to do so also means that you are not doing justice to the work you actually did over a span of several months. This article gives an outline of the report structure and an idea about the possible contents within several subsections of a BEng, MEng, MSc or PhD reports.

A good report generally follows the following order:

Title page
It should include tile of the report, name of the department and University with logo, few words explaining its submission (i.e. this dissertation is submitted for (or the part of) the degree of B.Eng., MEng or PhD, etc.), followed by your name and supervisors name (in some case people do not write it here, so its better to check in earlier reports).

Dedication (optional)
It’s an optional page, but you want to dedicate your report to someone by writing his/her name (i.e. dedicate to….)

Declaration
It is an important page, and should include a couple of lines stating that the work presented in the report is your own work and contents taken form other places are duly acknowledged. Also, a line including the information on number of words, tables and figures can be included.

Acknowledgements
Acknowledging someone does not affect you, but leave a good impression on the person who helped you. Firstly, one should clearly acknowledge his/her supervisor and co–supervisor (if any) with their names. Then, it is worth acknowledging others (including funding body, colleagues, family and friends) if you feel that they have contributed directly or indirectly to your work.

Publications
Include a page on list of publications (if you published any) that came out as a part of work included in your report. It is a direct measure of the quality of your work, giving a clear idea to the examiner about the originality of your work , providing you a strong chance to defend your work. For PhDs, it is quite important to list journal and conference publications. It will look impressive and extraordinary if one can tick this box while doing other degrees.

Abstract
Abstract is the ‘face’ of your report. It should be as concise and short as possible. A standard is generally of 350 words or one page length. If one is not being able to fit this within this criteria, it reflects his confusion and inability to summarise the key results in a consolidated manner.

An abstract should address the following questions in a orderly manner. Couple of lines can be dedicated to each of the following:
  • Motivation: why have you done this study or what are research gaps,
  • Objective: what are aims of your work,
  • Methodology adopted to address the define objectives,
  • Key results and findings,
  • Application of your findings to real–world problems or how this will fill the research gap.
Although, it is up to an individual that how much word he dedicate to each of the above questions.

Note that the pages before this (including abstract) are NOT numbered.

Table of Contents
This should include a detailed list of all the contents starting from this section to Appendices. Numbering should start from this page in roman letters (i.e. i, ii. iii, … etc.) and should end at List of nomenclature. From Chapter 1 (Introduction) real numbering of the report should start (i.e. 1, 2, 3….).

List of Figures

List of Tables

List of Nomenclature

Chapter 1: Introduction
It should contain the following: (i) motivation, (ii) research objective, (iii) research approach, and (iv) report outline. It is always good to keep this chapter short (i.e. 3-5 pages), but length can vary depending on the type of report.

Chapter 2: Literature Review

It should contain the literature review of studies/materials which you are going to use in your upcoming chapters. For a BEng, MSc or MEng report, it can be between 10 and 20 pages but these numbers can be higher for PhD reports.

Chapter 3: Materials and Methods

This chapter should contain the description of methods which you are using for your main research work. In case of computational or theoritical study, it can contain the methods developed or applied. Number of pages may vary depending on the type of report and work.

Chapter 4: Results and Discussions
These can have more than one chapter depending on the nature of the report. These are the main and most important chapters on which you will be evaluated; these chapters reflect your original contribution to the selected topic. Thesrefore, it is important that these chapters are written concisely in a simple and clear language.


There are few points which can be considered while writing these chpaters: (i) one should not live in their own world while presenting the work but is always recommended to compare their results with already published someone else's work, (ii) these chapters should not just report the results but should include discussions on the results presented - sometime it is better to raise questions yourself to raise curocity in readers mind and then address them, (iiii) in some cases students have too much figures or material to include but it is dangerous to include all of them without their proper discussion - therefore the best way is to filter out the results (or try to merge figures together, if possible) and include the
best figures/tables that fits well to the theme of the work, and (iv) it ihas often been observed that students get very impressed with someone else's work and bury their original contribution while discussing other's work - it is very important that you distinguish the difference between yours and others' work and present your work as main contents and use others' work to compliment or justify your results.

Number of pages (or word count) may differ depending on the type of the report. However, it is generally expected that this part contain bulk (~60% or more) of the total word count.

Chapter 5: Summary, Conclusions and Future Work
This chapter should summarise your whole work in 1-2 pages and make your conclusions using bullet points in further one page. One should not forget to include one or more paragraphs on future scope of work.

References

Besides related books that you can search from internet or library, you should look for research articles. Few of the search engines may be Science-direct or ISI Web of Knowledge or Scopus.

It is an important part of any report. You should provide a reference to any work which you refer (i.e. done by others). Moreover, the references in your report should be consistent in writing style. You can use EndNote or similar softwares for making the job easier.

Appendices

Include all of them in order at the end of the report (i.e. after references) but include only those contents that has been referred in the main body of the text.

One more point to remember:

Editing

Editing is crucial for a report. Without a proper editing a report is 'like a very expensive cloth stiched in a poorly manner'. Such reports do not leave good mark on the readers/ or examinors. This is not difficult to do proper editing. If you are using Latex, its automatically take care most of the things but you can also do good editing in word by defining style, headings, etc. One should always try to maintain consistency in writing, font type and size, spacings between (or after) paragraphs, figures and tables, and referencing styles, etc.

Happy reading!
Prashant